A subgroup of prostate cancer patients treated with brachytherapy experience radiation-induced injury manifested as either urinary morbidity, proctitis or erectile dysfunction (ED). Results have been obtained from a series of studies suggestive of a genetic basis for clinical radiosensitivity and it has been hypothesized that many patients who exhibit normal tissue radiation toxicity harbor specific single nucleotide polymorphisms (SNPs) and copy number polymorphisms (CNPs) associated with a susceptibility for the development of adverse effects resulting from a radiation treatment for prostate cancer. However, the research performed to date has been restricted to genotyping only a limited number of SNPs in a small group of candidate genes. In recognition of our inadequate understanding of the pathways involved in the development of radiation-induced urinary morbidity, proctitis and ED, as well as the incomplete knowledge of the spectrum of genes/proteins involved in the development of these forms of radiation toxicity, it is likely that we have failed to identify many of the SNPs and CNPs that are associated with the development of these manifestations of radiation injury. Therefore, we are proposing a new and innovative strategy to achieve this goal in which a genome wide association study will be performed to discover a more complete spectrum of the SNPs and CNPs (and genes) that are associated with clinical radiosensitivity. This will be a case-control study in which each prostate cancer patient that develops either urinary morbidity, proctitis or ED will be matched on age, race, stage, date of diagnosis and dosimetric parameters, with an appropriate control patient who did not develop that form of radiation injury. There will be 200 cases and 200 controls for each form of radiation injury in this study. Half of the subjects will first be screened for SNPs and CNPs using the Affymetrix 6.0 SNP array. The type I error (a) for rejection of the null hypothesis will be set at 0.0001. Therefore, depending on the minor allele frequency, we will be able to identify SNPs and CNPs whose genome relative risks (GRRs) for the development of each form of radiation induced injury is greater than approximately 2.5. Although this is a relatively modest number of subjects for a genome wide association study, therefore enabling identification of SNPs or CNPs only with relatively high GRRs, it is important to note that only SNPs and CNPs with GRRs greater than roughly 2.5 will likely be of useful predictive value in the actual clinical setting considering the dosimetric uncertainties associated with a standard radiotherapy treatment. A second phase validation study will be performed with a separate replication set comprising the other half of the subjects selected for this project using an a of 0.01, which should eliminate virtually all false positives identified in the initial phase. Finally, we will perform comprehensive SNP screening for all subjects of the DNA region surrounding every SNP that proves positively associated with each form of radiation injury in the replication set of subjects in order to genotype all SNPs in a haplotype block.